Issues in semantic memory : a response to Glass and Holyoak
Edward Joseph Shoben · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign) · 1978
Glass and Holyoak (1975) have raised two issues related to the distinction between set-theoretic and network theories of semantic memory.They contend that:(a) their version of a network theory, the Marker Search model, is conceptually and empirically superior to our version of a set-theoretic theory, the Feature Comparison model; and (b) the contrast between set-theoretic and network theories parallels distinctions in formal semantics that are concerned with analyticity and binary truth values.We take issue with both of these claims.We first argue that the set-theoretic vs. network distinction is orthogonal to issues like analyticity and binary truth.Then we take up the Marker Search and Feature Comparison models in detail.We raise objections to some of the theoretical mechanisms postulated in the Marker Search model, and then deal with Glass and Holyoak's criticisms of the Feature Comparison model.Lastly, we present new experimental results that undermine the critical empirical base for the Marker Search model.